The Viral Epidemiology Branch (VEB) has used a variety of approaches to define the nature and magnitude of malignancies associated with HIV-1 and other chronic infections, including analyses of population-based data, prospective cohort studies, and laboratory investigations. We are currently conducting a limited follow-up of 2,480 subjects with CDC-defined AIDS (i.e., <200 CD4+ T-cells/ul) recruited in 1998-99. Analyses of this cohort have identified alcohol drinking as a risk factor for antibody reactivity to HHV-8, the Kaposi's sarcoma (KS)-associated herpesvirus. Among HHV-8 seropositives, alcohol was additionally associated with increased risk of KS. Similarly, cigarette smoking was protective both against HHV-8 antibody as well as against KS risk among HHV-8 seropositives. Alcohol and cigarettes each appear to have consistent effects at multiple stages of KS etiology, and further studies are planned to identify the underlying mechanisms. Nested case-control studies of incident cancers are planned, utilizing the detailed clinical and interview data and banked blood specimens (plasma and cryopreserved lymphocytes) collected at enrollment. We are currently planning an intensive laboratory investigation of non-Hodgkin's lymphoma (NHL)-associated mutations in these lymphocyte samples and their potential relationship to risk of NHL. Findings will be compared for HIV-related and unrelated NHL. Our two previous cohort studies of HIV-infected hemophilia patients have been incorporated into a larger cohort which includes HIV uninfecteds, whose prospective followup for cancer has recently been initiated. These cohorts will be applied to studies of potential risk factors for AIDS-related malignancy, including genetic polymorphisms and/or altered baseline levels of cytokines and other immune factors, chromosomal translocations, EBV and HHV-8 viral loads, antibody titers, and cell mediated immunity, and environmental exposures and medications. We are continuing our studies of EBV infection and immune response as a causative factor in AIDS-related lymphoma, with determinations of viral load, cell-mediated immunity, antibody reactivity, and other immune system parameters. These studies also include examination of host genetic variation in B-cell-stimulatory cytokines. In collaboration with the Occupational Epidemiology Branch, we are continuing our investigations of cytokine polymorphisms as risk factors for gastric cancer. In a collaboration with the Veterans' Administration, we are investigating a cohort of liver disease patients originally recruited in the 1970s, for which we are assessing effects of baseline hepatitis C viremia and subsequent infection with HIV.